摘要:Osiecki, R, Nascimento, VB, Daros, LB, Becker, D, Stanganélli, LCR, Dourado, AC. Prediction of Swimming Performance Over 200 m and 400 m Freestyle through the Log-Log Model and Critical Power Model. JEPonline 2014;17(1):58-66. The purpose of this study was to compare the prediction of swimming performance in 200 m and 400 m tests through the Log-Log Model and the Critical Power Model. Eight male swimmers performed maximal crawl swim performance in a 25 m pool over distances of 50 m, 100 m, 200 m and 400 m. The Log-Log Model showed better agreement (bias = - 0.23 sec, P = 0.456; r = 0.399, P = 0.326), compared to the Critical Power Model (bias = - 5.52 sec, P = 0.03), for real-time 200 m crawl when using predictive distances 50 m and 100 m. As to the distance of 400 m, the Log-Log Model, using the predictive distances of 100 m and 200 m, showed a better agreement (bias = - 2.5 sec, P = 0.382; r = 0.617, P = 0.102) than the Critical Power Model (bias = -14.25 sec, P = 0.003, r = 0.690, P = 0.058) when compared to real time. The findings indicate that the Log-Log Model is a better predictor of performance for distances of 200 m and 400 m freestyle (crawl swimming) than the Critical Power Model. Hence, it can be used to plan test strategies and to design swim training.